The regulatory variation is believed to play an important role in shaping phenotypic differences among individuals and thus is also very likely to influence disease susceptibility and progression. In this study, we propose to take advantage of the expression QTL mapping and co-expressed gene network analysis to identify and characterize candidate genes and genetic variants that are responsible for aggressive phenotype of prostate cancer. Our hypothesis is that most genetic variants responsible for an aggressive phenotype of prostate cancer have regulatory effect on candidate gene expression and complete understanding of regulatory SNPs can only be achieved by examining primary tissue (here, prostate). To test this hypothesis, we will use a case-case study design and apply an innovative yet feasible approach by integrating DNA sequence variation and gene expression with clinical trait information. The four Specific Aims are: 1. Identify novel aggressiveness-related candidate SNPs by utilizing an expression genetics-based eQTL mapping approach;2, Identify novel aggressiveness-related candidate SNPs by utilizing an integrative systems genetics-based network analysis approach;3. For the novel candidate SNPs identified in Aims 1 and 2, perform additional association-based studies to confirm their association with an aggressiveness-related phenotype for prostate cancer;and 4. Identify candidate causal-SNPs by fine mapping, recognizing that the candidate eSNPs identified in Aim 3 will most likely be in linkage disequilibrium with the causal-SNPs. Understanding genetic mechanisms underlying the aggressive phenotype will have significant impact on prevention strategies, prognosis and potentially targeted therapy.